The present study was separated into three experiments. First (Experiment 1), in order to validate the time point of termination after the last WBV session, the long-term effects (5 weeks) of WBV on cholinergic activity in the medial septum and hippocampus were studied in young mice. For this, mice were randomly assigned by a randomizer application to four experimental groups: two WBV-treated groups (n = 7 and 8) and two pseudo-WBV-treated groups (n = 7 and 8). After 5 weeks of WBV, a WBV and a pseudo-WBV control group were terminated 2 h after the last WBV session, and the other two groups (WBV and pseudo-WBV) were terminated 24 h after the last WBV session. This exploratory study was aimed at determining the temporal dynamics of WBV on cholinergic activity by investigating its effects at either 2 or 24 h after the last WBV session. This experimental design is shown in Fig. 1A.

Fig. 1.

Experimental design. This study consisted of three separated experiments. Gray blocks indicate days with a whole-body vibration (WBV) session of 10 min, and open blocks indicate days without WBV sessions. In Experiment 1 (A), 8-week-old male C57BL/6 mice were assigned to four experimental groups: two WBV-treated groups (n = 8/gp) and two pseudo-WBV-treated groups (n = 7/gp). Mice were terminated after 5 weeks of WBV. A WBV group + its corresponding pseudo-WBV group was terminated either 2 h or 24 h after the last WBV session (†). In Experiment 2 (B), 8-week-old male C57BL/6 mice were randomly distributed among 10 experimental groups: 5 groups underwent WBV and 5 corresponding groups served as home-cage controls (n = 8/gp). Mice were euthanised after 1, 2, 3, 4 or 5 weeks of WBV (24 h after the last WBV session). A baseline measure of motor coordination was obtained for mice in all groups before the start of intervention (indicated by *). Motor coordination of mice in all groups was reassessed at the end of the intervention, one or two days before termination (†). Brain tissue was collected for immunohistological purposes in the first and second experiments (†). In Experiment 3 (C), 8-week-old male C57BL/6 mice were randomly assigned to two experimental groups: a WBV treated group and a pseudo-WBV treated group (n = 13/gp). Spatial-reference memory was analyzed by a Y-maze reference-memory test at the end of 5 weeks of WBV intervention. The WBV intervention continued during the Y-maze test period, but was always done after the test session in order to avoid direct interference of the intervention with the test performance.

In Experiment 2, the long-term effects of WBV on cholinergic activity (and motor coordination) were studied with a longitudinal design. Mice were randomly assigned to 10 experimental groups (n = 8 each): 5 WBV-treated groups and 5 home-cage control groups. A balance-beam test was performed to assess motor coordination. A baseline measurement was obtained for all experimental groups at the start of the intervention, and a correspondent home-cage control group and a WBV group were tested every week before euthanasia. From the second week onwards, after each week, a WBV and a home cage control group were terminated 24 h after the last WBV session (Fig. 1B). This exploratory study was designed to investigate the longitudinal, weekly development (from week 1 to week 5) of ChAT-immunoreactivity in the hippocampus (dentate gyrus, CA3, and CA1). Next, based on the obtained ChAT results (see below in Section 3.2 Experiment 2), we analyzed synaptic density in either week 2 or 5 using synaptophysin as a marker. In addition, we knew from our previous study [20], that the body weight of mice was not affected by 5 weeks of WBV intervention with daily 10-min sessions. However, we did not examine this previously by the comparison of WBV mice and home-cage controls on a weekly basis. We weighed the mice every week in Experiment 2.

The long-term effects of WBV on spatial reference memory were investigated in Experiment 3. Mice were randomly divided into two experimental groups (n = 13/gp): a WBV-treated group and a pseudo-WBV group. The mice were tested in a Y-maze after 5 weeks of WBV to assess reference-memory function. The experimental design is shown in Fig. 1C.

One hundred and thirty-six C57BL/6 male mice were used in these experiments (Experiment 1, n = 30; Experiment 2, n = 80; Experiment 3, n = 26). The mice were 8 weeks old at the start of the WBV interventions. All mice were maintained in the same housing conditions: individually housed in cages (30 $\times$ 12 $\times$ 13 cm) containing nesting material, sawdust, and a Kleenex role. They were maintained on a 12-12 dark/light cycle (lights on at 0700 h and off at 1900 h), at a temperature of 21 $\pm$ 1 °C, and at a humidity of 50 $\pm$ 10%. They were given water and food ad libitum. The health of the mice was checked daily. In Experiment 2, their body weight was measured weekly. All experimental procedures were evaluated and approved by the national Central Authority for Scientific Procedures on Animals (CCD) and by the local Institutional Animal Welfare Body of University of Groningen (approval number: DEC5685).

We adhered to the reporting guidelines for WBV studies in animals [39]. Mice underwent the same WBV procedure in all three experiments, including a single, daily WBV session of 10 min, 5X/week. No habituation to the device before the start of the WBV protocol took place. The WBV device had been used by our research group earlier [20, 29, 40] and it consisted of a power amplifier (V406 Shaker Power Amplifier, Ling Dynamic Systems Ltd, Hertfordshire, UK) and an oscillator (LEVELL R.C. Oscillator Type TG200DMP, Levell electronics Ltd, Hertfordshire, UK) attached to a cage with 12 removable compartments (6.5 $\times$ 7.5 $\times$ 20 cm) (Fig. 2A–C). The parameters of mechanical vibration generated in the WBV device were set at a frequency of 30 Hz and vertical amplitude of 54 µm. These parameters have been previously tested and verified by additional measurements by an accelerometer (corner: Y: 60 µm, Y: 75 µm; Z: 54 µm; centre: Y: 40 µm, Y: 29 µm; Z: 14 µm). The weekly treatment schedule was randomised regarding the following: each group underwent WBV at a different time between 0900 h and 1100 h (mice were rotated to prevent potential time-of-day effects). During the WBV sessions, mice were randomly placed into one of the compartments and did not have social interaction with each other during the WBV treatments. These compartments were cleaned with 70% ethanol solution and dried with paper towels after each training session. All WBV sessions were conducted in the housing room during the light phase. It should be noted that mice showed slightly higher levels of locomotor behavior during the first week of intervention. However, from the second week on, they did not show additional unprompted activity and they mainly stayed in a sitting or lying position in the compartments. Further, it is important to note that we did not observe acute, short-term or long-term overt side effects of WBV. In the first and third experiments, a pseudo-WBV treatment served as the control. The mice underwent the same procedure (e.g., handling) but without actual vibration exposure. In contrast, in the second experiment we did not use pseudo-WBV groups because the aim was to determine the temporal dynamics of ChAT expression, for which we needed the most stable control situation (i.e., home-cage controls). Previous experiments showed WBV-specific changes (but not in pseudo-WBV controls), including ChAT-immunostaining in the amygdala [20, 29, 40].

Fig. 2.

WBV device. The whole-body vibration device used in this study consists of 3 main compartments (A): (1) the oscillator (panel B – upper device); (2) the power amplifier (B – lower device), and (3) the cage with 12 removable compartments (C). This system provides a constant frequency of 30 Hz and amplitude of 54 µm, and allows for the treatment of 12 mice simultaneously without physical interactions.

The balance-beam test was used to measure motor coordination in Experiment 2. It served as an outcome measure of muscle performance as well as an indicator of the functional impact of the WBV treatment (based on previous experiments [20, 40]). A 1-m long, 4-mm wide, wooden beam was placed horizontally 80 cm above a container filled with paper used for protecting the mice if they fall. A safe cage, which served as motivation for the mouse to cross the beam, was placed at the far end of the beam. Firstly, mice were allowed to familiarise themselves to the safe cage for 1 min. Secondly, the mice were placed in the middle of the beam to allow them to become familiar with the test environment. Finally, mice were placed at the near end of the beam facing the safe cage. The time to cross to the safe cage was recorded. Balance-beam sessions were carried out at the beginning of light phase.

A Y-maze task was used to assess spatial-reference memory functions, as described previously [41]. Briefly, the Y-maze task was carried out in a tubular, transparent, Plexiglas area consisting of a start arm and two test arms that formed the Y. The arms were positioned at 120° angles from each other and were 5 cm in diameter and 27.5 cm long. A start box was attached to the start arm. A food reward consisting of a small crumb of regular food (0.05–0.1 g) was placed in one of the test arms as bait. Food crumbs were also placed below the perforations at the end of the other test arm. This ensured that both arms would smell the same and thereby helps to avoid discrimination between the baited and non-baited arms based on smell. A small rim (1 cm high) was placed at the end of the arms to prevent visual inspection of the bait. A guillotine door was installed halfway down each arm. When the mouse chose an arm, the other arm was closed by the guillotine door, which was operated by the researcher.

Each mouse underwent two habituation trials on the first day to familiarise them with the experimental situation. In the first habituation trial, the mouse was placed in the start box, from which it was allowed to explore one of the two test arms, which was baited with a small crumb of food. The other arm was closed. After the mouse consumed the bait, it was allowed to roam back to the start box. The second habituation trial was immediately after the first, and the other test arm was opened and baited. After the habituation, mice performed a training protocol for seven days consisting of one training session/day. Each training session consisted of 6 trials. During the entire training phase, the role of the left–right test arms (baited or not baited) was randomly switched. A valid visit or choice was considered to have occurred when the mouse placed all of its paws in a test arm (and the experimenter closed the other arm). After the mouse returned to the start box, the start arm attached to the start box was closed, preventing the mouse from reentering the maze. After cleaning both the start and test arms and re-baiting the same arm, the test procedure was repeated again, allowing the mouse to re-explore the maze. A correct choice occurred when the mouse visited the baited arm. Correct choices for each training session were used as final outcome measures and determined the longitudinal learning curve of the mouse. In addition, the area under the curve (AUC) was calculated to describe the overall performance of the mouse in the Y-maze task. All Y-maze procedures were carried out at the beginning of the light phase.

Mice were terminated with CO₂ [42], directly followed by transcardial perfusion with saline (0.9% NaCl solution and 1 unit/mL heparin) (02405822, Apotheek Albert Schweitzer Ziekenhuis, Dordrecht, The Nederlands) and then with a 4% para-formaldehyde solution (158127, Sigma-Aldrich, St. Louis, MO, USA). Brains were collected and fixed in 4% para-formaldehyde solution for 24 h. After fixation, the brains were washed in phosphate buffer solution for 24 h, then were dehydrated in 30% sucrose solution. Samples were frozen with liquid nitrogen and stored at –80 °C until sectioning. Coronal sections of 20 µm were cut by cryostat, and free-floating sections were stored in phosphate-buffered solution containing 0.2% sodium azide at 4 °C until the immunohistochemical staining procedure.

The cholinergic marker ChAT (the rate-limiting enzyme for the production of acetylcholine) was visualised in both experiments with immunohistochemistry. Before the incubation of the primary antibody, sections were preincubated in 0.01 m Tris-buffered saline (TBS) containing 5% normal rabbit serum (NS01L, Sigma-Aldrich, St. Louis, MO, USA) and 0.2% Triton X-100 (TX) (X-100, Sigma-Aldrich, St. Louis, MO, USA). Primary antibody (Goat antibody ChAT, 1:400; Millipore Sigma, Burlington, NJ, USA) was incubated in 0.01 M TBS, 5% normal rabbit serum, and in 0.2% Triton X for 3 h in a 37 °C water bath, followed by 1 day at room temperature and 2 days in a cold room at 4 °C. Secondary antibody (Rabbit anti goat, 1:400; Jackson ImmunoResearch, Baltimore Pike, West Grove, PA, USA) was incubated in 0.01 M TBS containing 1% normal rabbit serum and 0.2% Triton X for 4 h at room temperature, then allowed to stand overnight in a cool room.

Synaptophysin, an integral membrane protein in the presynaptic vesicles of neurons, was stained to visualise presynaptic boutons and served as an indicator of synaptic density in Experiment 2. Sections were preincubated in 0.01 M TBS, 3% bovine serum albumin (A4503, Sigma-Aldrich, St. Louis, MO, USA), and 0.5% TX for 1 h. Primary antibody (mouse antibody, 1:1000; Millipore Sigma, Burlington, NJ, USA) was incubated in 3% bovine serum albumin and 0.5% TX in 0.01 TBS for 3 h at room temperature, followed by 3 days incubation at 4 °C. Secondary antibody (Goat anti mouse, 1:400; Jackson ImmunoResearch, Baltimore Pike, West Grove, PA, USA) was incubated in 0.01 TBS for 2 h at room temperature.

Before incubation of the primary antibodies, brain sections were prewashed in 0.01 M TBS followed by incubation of 0.3% H₂O₂ for 30 min. Finally, all sections were incubated with avidin-biotin peroxidase complex for 2 h (Vestastain ABC kit, Vector, Burlingame, CA, USA) and visualised with 0.075 mg/mL 3,3^′-diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA) solution. During the procedures, sections were extensively rinsed in 0.01 M TBS or in TBS containing 0.2% TX. Sections were mounted on microscopy slides, dehydrated in gradients of ethanol-xylol solutions and cover slipped for microscopic analyses.

Quantification of the immunohistochemical data was performed with a Quantimet system (Leica DFC 365FC camera, Leica Biosystems, Nussloch, Germany) integrated into a microscope system (Olympus BH2 microscope, Olympus-Life science, Shinjuku-ku, Tokyo, Japan) run by Quantimet software (Quantimet 570, Leica Biosystems, Nussloch, Germany). Field-size analysis was performed to ensure that Quantimet data collection was performed objectively. No significant difference in field sizes were found between experimental- and home-cage or pseudo-WBV control groups.

Optical density for ChAT was analyzed in the granular and molecular layers of dentate gyrus inner blade (DGI) for Experiment 1; and in both the DGI and outer blades (DGO) for Experiment 2. In addition, ChAT optical density was measured in the oriens and pyramidal layers of CA1 and CA3 subregions for both experiments, as well as in the medial septum (cell body and dendrites were separated) for Experiment 1.

The optical density for synaptophysin was obtained in the molecular layers of DGI and DGO blades for Experiment 2. These regions were selected because they were representative for the ChAT results and the high quality of the synaptophysin immunoreactivity. The obtained data were corrected for a specific background staining in the corpus callosum. Analysis was performed on pictures taken at 200$\times$ magnification.

All statistical analyses were done using Microsoft Excel (Microsoft Office Professional Plus 2019, Microsoft Cooperation, Redmond, WA, USA). Student’s independent t-tests were performed to evaluate the differences between the WBV and pseudo-WBV or home-cage control groups. Outliers (more than 2 Standard Deviations (SD) from the mean) were removed. Significance was defined as p $\le$ 0.05.

In Experiment 1, we determined whether septal and hippocampal ChAT immunoreactivity was altered after 5 weeks of WBV. In addition, we determined whether the time of termination (either 2 or 24 h after the last WBV session) influenced the outcome. Hippocampal ChAT expression was measured in subregions of the dorsal hippocampus (DGI, CA3, and CA1) and in the medial septum. ChAT-immunoreactivity visualised the cholinergic fibers arising from the cholinergic cells of the medial septum that innervated the different hippocampal subregions.

ChAT-positive immunostaining was not changed in the cholinergic cell bodies after WBV, either at the 2-h hour or the 24-h time point. In contrast, a significant increase in the axons and dendrites of the medial septum was found in the mice terminated 2 h after the last WBV session (p 0.05) (Fig. 3A). This was not observed at 24 h after the last WBV session. In the hippocampus, changes in the CA1 and CA3 were predominantly found in the pyramidal cell layer. A significant increase of the level of ChAT was detected in this layer in the 24-h termination group (p 0.05) (Fig. 3B,C). In addition, a significant increase of ChAT expression was also observed in the granular layer of DGI both in the 2-h and the 24-h groups (Fig. 3D). The same significant increment of ChAT immunoreactivity was found in the molecular layer of DGI in the 2-h group (p 0.05). An increase was observed in the 24-h group, but it was not statistically significant. Taken together, the septum responded with an increase in ChAT-immunostaining at the 2-h time point, as did the dentate gyrus (DG). Notably the CA1 and CA3 pyramidal layers responded after 24 h, as did the DGI in the granular layer. This leads to the general impression that the medial septum, the region of ChAT origin, showed the impact of WBV sooner than did the hippocampus, the main target region of ChAT that is transported by the cholinergic fibers to the terminals.

Fig. 3.

Experiment 1—ChAT immunostaining in the medial septum and parts of the hippocampus. Effects of 5 weeks of whole-body vibrations (WBV) intervention on the progression of choline acetyltransferase (ChAT) immunostaining in mice terminated after either 2 or 24 h after the last WBV session in the cell bodies and dendrites of medial septum (A), three layers of Cornu Ammonis 1 (CA1) (B), and CA3 (C), as well as in the granular and molecular layers of the dentate gyrus inner blade (DGI) (D). Cholinergic activity was significantly higher in the dendrites and axons of the medial septum in the 2-h group (A). WBV significantly increased ChAT levels in the pyramidal layer of CA1 (B) and CA3 (C) subregions in the 24-h group. ChAT level was significantly higher in the granular layer of DGI both in the 2- and 24-h groups (D). The molecular layer of DGI was only significantly affected by WBV in the 2-h group. Data are presented as mean $\pm$ standard error of the mean (SEM). * indicates p 0.05.

In Experiment 2, we investigated the longitudinal profile of ChAT immunoreactivity and motor coordination after 1, 2, 3, 4, or 5 weeks of WBV. In order to obtain comparable data of a possible longitudinal ChAT fluctuation induced by WBV over 5 weeks, as well as exclude the potential effects of other factors on ChAT level such as daily handling, stress, or novel environment, undisturbed mice served as home-cage controls.

Mice were weighted every week to assess possible effects of WBV on body weight. A daily 10-min session of WBV, 5 days/week, for 1–5 weeks of treatment had no significant effect on body weight. A balance-beam test was performed for evaluation of motor coordination in this second experiment. Balance-beam performance was not significantly different between the WBV and home-cage controls before treatment or after the first or second week of WBV. A significantly shorter walking time (indicating better motor coordination and performance) was found in the WBV-treated group than in the home-cage group from the third week onward (weeks 3 and 4: p 0.05; week 5: p 0.001) (Fig. 4). Motor coordination was most strongly affected after five weeks of WBV.

Fig. 4.

Experiment 2—Walking time on the balance beam as measure of motor coordination. Effects of 1–5 weeks of whole-body vibrations (WBV) intervention on progression of motor coordination in the balance-beam test, showing the functional impact of WBV. Balance-beam performance was significantly improved from week 3 onward by the WBV intervention. Data are presented as mean $\pm$ SEM. * indicates p 0.05.

We further scrutinised the impact of WBV on ChAT immunoreactivity in Experiment 2. To determine the dynamics of ChAT levels over the 5-week intervention period, we measured the optical density of ChAT staining in subregions of the dorsal hippocampus (DGI, DGO, CA3, and CA1) at either week 1, 2, 3, 4 or 5. Home-cage mice served as controls to provide the most stable ChAT level for comparison. The home-cage controls showed a significantly higher level of ChAT-positive staining after 2 weeks of the experiment than did the WBV-treated group. This significant difference was detected in the stratum oriens of CA1 and CA3 (CA1: p 0.01; CA3: p 0.05) (Fig. 5A,C) and in the pyramidal cell layer of CA1 (p 0.01) (Fig. 5B), as well as in all subregions of the DGI and DGO blades (p 0.05) (Fig. 6A–D). In contrast, the WBV-treated group had a significantly higher ChAT level after 5 weeks of WBV than did the home-cage control group after 5 weeks. This difference was statistically significant in all subregions (CA1: p 0.01; CA3: p 0.05; DGI: p 0.01; DGO: p 0.05) (Fig. 5A–D and Fig. 6A,B,D), except in the granular layer of DGO blade (Fig. 6C). ChAT-immunoreactivity in the WBV group and the home-cage controls did not differ significantly on weeks 1, 3, and 4. It is important to stress that Experiment 1 indicated that the increase in ChAT levels in the stratum oriens most likely were not WBV-specific (and hence must be a result of the general handling and behavioral activation of the mice). In contrast, experiment 1 suggests that the observed changes in the pyramidal cell layer and the DGI in Experiment 2 are indeed WBV-specific. Images of Chat immunostaining were taken from the CA1 and DGI blade subregions (Fig. 7A). These representative images are shown in Fig. 7C,D.

Fig. 5.

Experiment 2—ChAT immunostaining in the hippocampal CA1 and CA3 subregions. Effects of 1–5 weeks of whole-body vibration (WBV) intervention on the progression of choline acetyltransferase (ChAT) immunostaining in the CA1, stratum Oriens (A), and pyramidal layer (B); and in the CA3, stratum oriens (C), and pyramidal layer (D). ChAT levels were significantly lower in the WBV group after 2 weeks and higher in the WBV group after 5 weeks. All data are presented relative to the baseline level (100%) of the home-cage control groups. Data are presented as mean $\pm$ SEM. * indicates p 0.05; ** indicates p 0.01.

Fig. 6.

Experiment 2—ChAT and synaptophysin immunostining in the hippocampal dentate gyrus subregions. Effects of 2 or 5 weeks of whole-body vibration (WBV) intervention on the progression of choline acetyltransferase (ChAT) immunostaining in the granular (A) and molecular layers (B) of the dentate gyrus inner (DGI) blade, as well as in the granular (C) and molecular layers (D) of dentate gyrus outer (DGO) blade. Bottom panels: the effects on synaptophysin immunostaining in the molecular layer of DGI (E) and DGO blades (F). ChAT levels were significantly lower in the WBV group after 2 weeks; ChAT levels were significantly higher in the WBV group after 5 weeks. Level of synaptophysin was only higher in the DGO blade after 5 weeks of WBV. All data are presented relative to the baseline level (100%) of home-cage control groups. Data are presented as mean $\pm$ SEM. * indicates p 0.05; ** indicates p 0.01.

Fig. 7.

Experiment 2—Representative pictures of ChAT and synaptophysin immunostaining in selected hippocampal regions. Schematic overview of hippocampal regions of interests for choline acetyltransferase (ChAT) and synaptophysin (A). Representative images of the synaptophysin immunostaining in the dentate gyrus in weeks 2 and 5 are shown in (B). Representative images of ChAT immunostaining from week 1 to week 5 are visualised in the CA1 (C) and dentate gyrus inner (DGI) blade (D) hippocampal subregions. Images were taken at 200$\times$ magnification. Scale bars in B, C and D are 100 µm.

The aim of Experiment 2 was to analyze the longitudinal progression of ChAT immunostaining (as an indicator of cholinergic activity) induced by WBV. Quantification of the ChAT results showed significant differences in weeks 2 and 5. To examine whether a different marker would show a similar pattern, we performed an additional exploratory staining to assess synaptic density of two representative subregions: the molecular layers of DGI and DGO blades. Synaptophysin was stained and analyzed on either week 2 or 5 in the DGI and DGO subregions. Synaptophysin did not show significant alteration after 2 weeks of WBV despite a reduction in expression (Fig. 6E,F). In contrast, similarly to ChAT immunostaining, synaptophysin was significantly higher in the WBV-treated group after 5 weeks than in the home cage controls (DGO: p 0.05) (Fig. 6F). Representative images of synaptophysin immunostaining were taken in the DGI and DGO blades (Fig. 7A) and are visualized in Fig. 7B.

In Experiment 3, we investigated the effects of a 4-week-long WBV intervention on spatial reference memory functions in a Y-maze task. Spatial memory performance was not significantly different in the WBV and pseudo-WBV control groups in the first and last sessions of the Y-maze task. However, significantly more correct choices were made by the WBV-treated group than by the pseudo WBV-treated group from the second session to the sixth session (session 2–6 p 0.05) (Fig. 8A). The number of correct choices was especially higher in the third session. Furthermore, the calculated AUC was significantly higher in the WBV group than in the pseudo-WBV group (AUC: p $\le$ 0.001) (Fig. 8B).

Fig. 8.

Experiment 3—Spatial reference memory performance in the Y-maze. Effects of 5 weeks of whole-body vibration (WBV) intervention on spatial reference memory functions in the Y-maze task. WBV-treated mice made significantly more correct choices from session 2 to session 6 than did the pseudo WBV-treated controls (session 2–6: p 0.05) (A). The area under the curve was calculated to represent an overall outcome measure of all Y-maze task sessions (B). Mice subjected to WBV showed a significantly higher value of the area under the curve (AUC: p 0.001). Data are presented as mean $\pm$ SEM. * indicates p 0.05; *** indicates p 0.001.